Sturdivant and Clark: Effects of Callmectes sapidus behavior on the efficacy of crab pots for estimating population abundance 
51 
was chosen to simulate high density conditions (Clark 
et ah, 1999). Crabs were placed in the mesocosm an 
hour before the start of the experiment and allowed to 
acclimate. An hour after acclimation, the CTV camera 
system was inserted into the center of the mesocosm. 
At the end of the 24-h experiment, the CTV cam- 
era was removed, and the number of crabs caught was 
recorded. A new set of 6 large and 10 smaller male 
blue crabs were obtained for the next trial, and the 
procedure was repeated. All video recordings from the 
experiments were analyzed at SERC. The number of ap- 
proaches, entries, escapes, and catch rates were record- 
ed, as well as behavioral interactions between crabs. 
Crab behaviors were classified into three qualitative 
categories: aggressive, agonistic, or neutral. Aggres- 
sive interactions were characterized by the extension 
of both chelipeds, and cheliped embracing or grasping. 
Neutral interactions were defined as those where the 
chelipeds were in a resting position while the crabs 
passed within 3.8 cm (the diameter of a mesh ring) of 
each other (Jachowski, 1974). Agonistic interactions 
comprised any other interactions that occurred, such 
as shielding (using the cheliped as a shield), fending off 
predators, poking, leaning backward, or leaning to the 
side (Jachowski, 1974). Only one crab needed to exhibit 
an aggressive or agonistic act for the interaction to be 
recorded as such. If an aggressive and agonistic act 
co-occurred, the interaction was defined as aggressive. 
Results 
Field experiment 
A total of 119 crabs were caught in 45 experimental runs 
for an average catch rate of 2.7 crabs per deployment. 
Crabs ranged in size from 81 to 179 mm CW (mean of 
142 mm [SE ±1.8]). Size of test crabs had no significant 
effect on the size of crabs caught, nor was there a signifi- 
cant size by depth interaction (ANOVA, P>0.05, F=0.63, 
df=4). There was a significant effect of depth (Fig. 2 A) on 
the size of crabs caught. Crabs caught at the 3-m depth 
were significantly smaller then crabs caught at 1 and 2 m 
(Tukey, P=0.03, F= 3.72, df=4). 
The size of test crabs had no significant effect on 
the number of crabs caught nor was there a signifi- 
cant depth-by-size interaction (ANOVA, P>0.05, F=0.11 
df=4). There was a significant effect of depth on the 
quantity of crabs caught (Fig. 2B). At the 1-m and 2-m 
depths, the number of crabs caught did not significantly 
differ. The number of crabs caught at 3 m was sig- 
nificantly less than at the 1-m and 2-m depths (Tukey, 
P=0.04, F=3.60, df=4). 
It is possible that the experimental design impacted 
the effect of the test crabs in our field experiment. In 
the field study, the test crabs were not tethered to the 
pot, therefore the possibility of escape existed. However, 
although the majority of experiments retained their 
test crab (-70%), if a test crab escaped from the pot 
before interacting with a conspecific, the pot essentially 
Figure 2 
(A) Mean size (±1 standard error [SE]) 
of crabs caught in relation to depth. 
Depth had a significant effect on size 
of crabs caught (P=0.03, F= 3.72, df=4). 
Pots at the 3-m depth caught signifi- 
cantly smaller crabs than pots at the 
1- and 2-m depths. (B) Mean abundance 
(±1 SE) of crabs caught in relation to 
depth. Depth had a significant effect 
on the number of crabs caught (P=0.04, 
F=3.60, df=4). Pots at the 3-m depth 
caught significantly fewer crabs than 
pots at the 1- and 2-m depths. Differ- 
ent letters denote significance. n = 15 
for each of the three depth treatments. 
Results were based on an analysis of 
119 crabs. 
became a control pot. The opposite held true for control 
pots. Once a crab entered a control pot, the control pot 
basically became a test pot because it then harbored a 
single crab. 
Of the crabs caught in the first 24 hours of the field 
experiment, 41% escaped before the end of the second 
